context.logger() – Contextual Logger API Reference

The context.logger() function returns a Logger instance automatically bound to the current run, node, and, when available, graph context. This logger provides the standard Python logging interface, while automatically injecting structured fields such as run_id, node_id, and graph_id into every log record.

Configuration:
Set logging options when starting the sidecar:

from aethergraph import start_server

start_server(log_level="warning", ...)

The runtime also respects environment and application settings (e.g., JSON vs. text output, log rotation, per-namespace log levels).

---

Log Record Fields

Each log record includes:

• run_id, node_id, graph_id (when available)
• Logger namespace (e.g., aethergraph.node.<id>)
• Message, level, timestamp
• Optional metrics or custom fields via extra={...}

---

Quick Reference: Logger Methods

Use the logger as you would a standard Python Logger:

Method	Purpose	Inputs	Notes
debug(msg, *args, **kwargs)	Verbose diagnostics	msg: str, *args, **kwargs	For high-volume internal logs.
info(msg, *args, **kwargs)	General information	msg: str, *args, **kwargs	Default level; typically visible.
warning(msg, *args, **kwargs)	Non-fatal issues	msg: str, *args, **kwargs	Alias: warn() (deprecated in stdlib).
error(msg, *args, **kwargs)	Recoverable errors	msg: str, *args, **kwargs	Use exc_info=True for tracebacks.
exception(msg, *args, **kwargs)	Error with traceback	msg: str, *args, **kwargs	Same as error(..., exc_info=True).
critical(msg, *args, **kwargs)	Severe/fatal conditions	msg: str, *args, **kwargs	May trigger alerts depending on sink.

Common kwargs:

• extra: dict — Attach additional structured fields (merged with contextual fields).
• exc_info: bool | BaseException — Include exception traceback.